Non commercial 'technical' publications often have useful tags and comments in their metadata, but incorrect title & author, and the original filename is often just a reference number. I'm thinking of public domain documents from academia, governments, think tanks, supra-nationals such as UN, IMF, EU etc. There are no 3rd party metadata sources for these publications.
For many people the 'title & author(s)' are the primary identification for their documents - so existing collections will often have file-names with accurate title and author, whilst the embedded metadata will remain untouched because without Calibre there's no convenient way to even view it - let alone get it right.
When I was initially loading Calibre with my existing collection of 'technical' publications I would have appreciated an option to take 'title & author(s)' from the file-name and the cover & other metadata from whatever was embedded in the document. I ended up adding books to a "New Books" library, taking all the metadata from the document, and then correcting the 'title & author(s)' before moving the books to the "Main" library. I still do this with all books - even real ebooks from Amazon, B&N etc.
Originally Posted by DazJWood
Incidentally, when I try to capitalize every word in the title (using Title Case), why does it not capitalize words like of, a, the, etc...?
"Title Case" like most things English, has no single authoritative definition see ===>>> http://grammar.about.com/od/grammarf...italstitle.htm
What you're asking for is more often called "Capital Case". What Calibre's Capitalize option does is more often referred to as "Sentence Case".
Other languages have different rules and conventions often maintained by government sponsored regulatory agencies - e.g. in French and German its probably defined by the Académie française
and Rat für deutsche Rechtschreibung